Learning Utilities from Demonstrations in Markov Decision Processes
This work addresses the issue of model misspecification in Inverse Reinforcement Learning for risk-sensitive applications, offering a novel approach to infer agent risk attitudes from demonstrations.
The paper tackles the problem of extracting risk attitudes from demonstrations in Markov Decision Processes by proposing a utility function model and defining the Utility Learning problem, resulting in two provably efficient algorithms with analyzed sample complexity and proof-of-concept experimental validation.
Our goal is to extract useful knowledge from demonstrations of behavior in sequential decision-making problems. Although it is well-known that humans commonly engage in risk-sensitive behaviors in the presence of stochasticity, most Inverse Reinforcement Learning (IRL) models assume a risk-neutral agent. Beyond introducing model misspecification, these models do not directly capture the risk attitude of the observed agent, which can be crucial in many applications. In this paper, we propose a novel model of behavior in Markov Decision Processes (MDPs) that explicitly represents the agent's risk attitude through a utility function. We then define the Utility Learning (UL) problem as the task of inferring the observed agent's risk attitude, encoded via a utility function, from demonstrations in MDPs, and we analyze the partial identifiability of the agent's utility. Furthermore, we devise two provably efficient algorithms for UL in a finite-data regime, and we analyze their sample complexity. We conclude with proof-of-concept experiments that empirically validate both our model and our algorithms.